JP6395640B2 - Method and apparatus and computer program for canceling narrowband interference in a single carrier signal - Google Patents

Description

  The present invention generally relates to a method and apparatus for canceling narrowband interference in a single carrier signal representing a received symbol.

  The present invention relates to canceling a narrow band interferer in a communication system based on single carrier modulation where demodulation is implemented in the frequency domain.

  By way of example and in a non-limiting manner, the present invention is applicable to single carrier orthogonal frequency division multiplexing (SC-OFDM).

  SC-OFDM is a modulation scheme that involves OFDM type multiplexing but has an envelope like a single carrier. This can be implemented both in the time domain and in the frequency domain. The latter case is also referred to as (discrete Fourier transform) DFT spread OFDM or SC-FDE (single carrier frequency domain equalization) or SC-FDMA (single carrier frequency division multiple access). Frequency domain implementation is generally preferred (especially in receivers).

  The present invention is also applicable to single carrier time division multiplexing (SC-TDM) when equalization is performed in the frequency domain.

  The present invention relates to a wireless cellular telecommunication network (such as 3GPP / LTE uplink transmission) or broadcasting system (Digital Video Broadcasting Next Generation Handheld (DVB-NGH) system and satellite communication system. Find an application).

  It is an object of the present invention to provide a method and apparatus that enables cancellation of at least one narrowband interference in a single carrier signal.

For this purpose, the present invention is a method for canceling narrowband interference in a single carrier signal, comprising:
The method comprises the steps performed by the receiver:
-Receiving a single carrier signal and converting the single carrier signal into received symbols;
-Converting the received symbols from time domain to frequency domain to receive symbols in the frequency domain;
-Determining a signal and thermal noise power estimate based on the received symbol power in the frequency domain;
-Estimating the variance of the narrowband interference from the signal and thermal noise power estimation and the received symbol power in the frequency domain;
-Taking into account the estimation of the variance of the narrowband interference, equalizing received symbols in the frequency domain or symbols derived from received symbols in the frequency domain, How to do.

Further, the present invention is an apparatus for canceling narrowband interference in a single carrier signal,
The device is included in a receiver;
The device is
-Means for receiving a single carrier signal and converting said single carrier signal into received symbols;
-Means for converting said received symbols from time domain to frequency domain to be received symbols in the frequency domain;
-Means for determining a signal and thermal noise power estimate based on the received symbol power in the frequency domain;
Means for estimating the variance of the narrowband interference from the signal and thermal noise power estimation and the received symbol power in the frequency domain;
-Means for equalizing received symbols in the frequency domain or symbols derived from received symbols in the frequency domain, taking into account the estimation of the variance of the narrowband interference It also relates to the device.

  In this way, the amount of interference in the received symbol is estimated for each frequency (ie for each carrier index), and equalization is improved when interference is present, and the overall performance of the receiver is improved when interference is present. To do.

According to certain features, the receiver
Determining a weighting factor that depends on the variance of the narrowband interference, wherein the weighting factor is a decreasing function of the variance of the narrowband interference;
Equalize received symbols in the frequency domain taking into account the weighting factor.

  Thus, equalization uses information about interference in a simple manner and reduces complexity.

  According to a particular feature, the weighting factor is equal to 1 or a null value, which depends on the estimated variance of the narrowband interference.

  In this way, the complexity of equalization is further reduced.

According to certain features, the receiver
-Estimate the frequency-dependent received power of received symbols in the frequency domain,
-Iteratively determining the signal and thermal noise power from the estimated frequency dependent received power.

  In this way, signal and thermal noise power can be estimated simply and efficiently, complexity remains low, and performance improves.

According to a particular feature, the adaptive signal and thermal noise power determined iteratively are:
Performing a first averaging process of the total received power of the received symbols in the frequency domain;
-Determining a threshold based on said average total received power in a first iteration;
Rounding all received symbol power in the frequency domain greater than the threshold determined in the first iteration;
Performing a second averaging of the rounded power;
-Correcting the second average by a correction factor;
-Determining subsequent adaptive thresholds based on the corrected average in subsequent iterations;
-Rounding off any power greater than the subsequent adaptive threshold;
-Performing a third averaging of the rounded power;
-Correcting the third average by a correction factor;
It is determined by executing the adaptive threshold determination, the rounding, the third averaging process and the correction a predetermined number of times.

  In this way, the signal and thermal noise power, which eliminates the interference power, is estimated easily and efficiently, and the correction factor ensures that the signal and noise power are not underestimated in the absence of interference, and the complexity remains low. The performance improves when there is interference.

  According to a particular feature, the receiver determines a first threshold based on the signal and the thermal noise power, and the first threshold is used to estimate the variance of the narrowband interference.

  In this way, the variance is easily estimated and the complexity remains low.

  According to a particular feature, the correction factor is calculated assuming that the symbol with rounded power follows a complex Gaussian law.

  Thus, in each iteration, the power loss due to rounding is compensated by the correction factor when there is no interference, and no degradation occurs when there is no interference.

  According to a particular feature, the correction factor is determined using a look-up table.

  In this way, the correction factor is easily determined without additional calculations.

  According to a particular feature, the receiver performs channel estimation based on received symbols in the frequency domain.

  In this way, the channel is estimated and equalization can use this estimate to improve the performance of the receiver.

According to certain features, the receiver
-Determining a second threshold based on said signal and thermal noise power estimate;
The rounded received symbol in the frequency domain is a symbol derived from the received symbol by rounding the amplitude of the received symbol in the frequency domain at the determined second threshold.

  In this way, the amount of interference in received symbols representing data and / or pilot symbols is reduced.

  According to a particular feature, the channel estimation is performed on the received symbols in the rounded frequency domain.

  In this way, the channel estimation performance is improved and the overall performance of the receiver is improved.

  According to a particular feature, the single carrier signal is a single carrier orthogonal frequency division multiplex modulation signal.

  In this way, the dedicated header or prefix facilitates demodulation in the frequency domain.

  According to yet another aspect, the present invention is a computer program that can be directly loaded into a programmable device, wherein the computer program is executed in a programmable device in a method according to the present invention. The present invention relates to a computer program including code portions or instructions for performing each step.

  Since the features and advantages relating to the computer programs are the same as those set out above related to the corresponding method and apparatus according to the invention described above, they will not be repeated here.

  The features of the present invention will appear more clearly from a reading of the following description of example embodiments made with reference to the accompanying drawings.

Fig. 2 represents a radio link in which the present invention is implemented. Fig. 2 is a diagram representing the architecture of a receiver in which the present invention is implemented. FIG. 3 is a block diagram of components of a radio interface of a receiver according to the first mode of realization of the present invention. FIG. 4 is a block diagram of components of a radio interface of a receiver according to a second mode of realization of the present invention. It is a figure showing the example of the received signal accompanying narrowband interference. FIG. 3 is a block diagram of a signal and thermal noise power estimation module according to the present invention. It is a figure showing the example of the algorithm performed by the destination by the 1st implementation mode of this invention. It is a figure showing the example of the algorithm performed by the destination by the 2nd implementation mode of this invention.

  FIG. 1 represents a radio link in which the present invention is implemented.

  The invention is disclosed in an example in which a signal transferred by a source Src is transferred to at least one receiver Rec.

  Although only one receiver Rec is shown in FIG. 1 for simplicity, it is also possible to receive signals by a more important number of receivers Rec.

  The receiver Rec may be included in a fixed terminal or a mobile terminal that is a destination to which data such as a video signal is transferred.

  Data (and possibly information that enables estimation of the radio link between the source and one receiver) is transferred using single carrier modulation.

According to the invention, the receiver Rec is
-A single carrier signal is received and a single carrier signal is converted into a received symbol.
-Convert received symbols from time domain to frequency domain to receive symbols in frequency domain,
-Determine signal and thermal noise power estimates based on received symbol power in the frequency domain,
-Estimate the variance of narrowband interference from signal and thermal noise power estimates and received symbol power in the frequency domain,
Taking into account the estimation of the variance of narrowband interference, equalizing received symbols in the frequency domain or symbols derived from received symbols in the frequency domain;

  FIG. 2 is a diagram representing the architecture of a receiver in which the present invention is implemented.

  The receiver Rec has, for example, an architecture based on components connected to each other by a bus 201 and a processor 200 controlled by the program disclosed in FIGS. 6a and 6b.

  It should be noted that the receiver Rec may have an architecture based on a dedicated integrated circuit.

  Bus 201 links processor 200 to read only memory ROM 202, random access memory 203, and to wireless interface 205.

  Memory 203 includes registers intended to contain variables and program instructions relating to the algorithm disclosed in FIG. 6a or 6b.

  The processor 200 controls the operation of the wireless interface 205.

  The read-only memory 202 includes program instructions relating to the algorithm disclosed in FIG. 6a or 6b, which are transferred to the random access memory 203 when the receiver Rec is powered on.

  Any and all steps of the algorithm described below in connection with FIG. 6a or 6b are instructions by a programmable computer (such as a PC (personal computer), DSP (digital signal processor) or microcontroller)). Or may be implemented in software by execution of a program. Alternatively, it may be implemented in hardware by a machine or a dedicated component (such as an FPGA (Field Programmable Gate Array) or ASIC (Application Specific Integrated Circuit)).

  In other words, the receiver Rec includes a circuit (or a device that includes the circuit), which causes the receiver Rec to perform the steps of the algorithm described below in connection with FIG. 6a or 6b.

  Such a device including circuitry that causes the receiver Rec to perform the steps of the algorithm described below in connection with FIG. 6a or 6b may be an external device connectable to the receiver Rec.

  The wireless interface 205 comprises the components disclosed in FIGS. 3a and 3b.

  Fig. 3a discloses a block diagram of the components of the radio interface of the receiver according to the first mode of realization of the present invention.

  The wireless interface 205 includes a synchronization module 301. The synchronization module 301 is responsible for synchronizing the DFT module 300 of the wireless interface 205 with respect to the received symbols.

The DFT module 300 converts the received symbol from the time domain to the frequency domain, and converts it into a received symbol y _k in the frequency domain. Here, k represents a carrier index. Received symbols are obtained by converting received single carrier signals into received symbols.

The received symbol in the frequency domain is
y _k = h _k x _k + ν _k
May be represented by Where h _k is the channel response for the carrier with index k and ν _k is the additive noise at the same frequency. The term ν _k is an additive white Gaussian noise (AWGN) addition, for example thermal noise and narrowband interference sources. For narrowband interferers, the dispersion of ν _k is frequency dependent and is represented by σ _k ² .

  Received symbols in the frequency domain are provided to adaptive signal and thermal noise power estimation module 305, threshold-based interference estimation module 303, channel estimation module 302, and equalization module 306.

Adaptive signal and thermal noise power estimation module 305 provides a signal and thermal noise power estimate (denoted P _s ) to threshold-based interference estimation module 303.

For example, the threshold-based interference estimation module 303 estimates the coarse interference source variance ω _k used by the equalization module for each carrier at index k. This dispersion is calculated from the received power p _k, the threshold T _a as calculated from the signal and the thermal noise power estimation P _s.

For example, T _a = P _s + xdB, and x is 0 to several dB (such as 10 dB at the maximum).

For example, ω _k = max (0, p _k −T _a ).

  This information representing narrowband interference is provided to equalization module 306.

The coarse dispersion ω _k is used by an equalization module 306 that performs equalization in various ways.

  The equalization module 306 may be minimum mean square error equalization (MMSE).

For example, the coarse variance ω _k is used in MMSE equalization as follows:

Where σ ² is an estimate of thermal noise variance and h _k ^* represents the conjugate of the estimated channel to carrier k.

Here, it should be noted that σ ² may be set up to a predetermined value.

For example, the coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

  If the equalization is MMSE equalization,

Here, it should be noted that λ _k is a decreasing function of ω _k . For example, λ _k is

Where, for example, a = b = 2P _s .

For example, lambda _k, depending on the value of omega _k, (or, where it is equivalent, depending on the relative values of p _k with respect to the adaptive threshold T _a) 0 and 1 of two Take only the value.

λ _k = 0 corresponds to the position of the narrowband interferer, and the equalization symbols on these carriers are set to null values.

  In accordance with the present invention, narrowband interference is reduced by providing information related to interference prior to equalization (eg, but not limited to Minimum Mean Square Error (MMSE) equalization). When present, the overall performance of reception is improved.

For each carrier k, the output omega _k or lambda _k thresholds based interference estimation module 303 is provided to the equalization module 306.

  Channel estimation module 302 performs channel estimation (eg, based on pilot symbols).

  The output of the equalization module 306 is provided to an IDFT (Inverse Discrete Fourier Transform) module 307. The IDFT module 307 may have a different size than the DFT module 300.

  Classical equalization (especially MMSE) assumes complete knowledge of the channel and narrowband interference source dispersion. Narrowband interference variance is easily estimated by the present invention. However, the channel estimation process is sensitive to interference sources. It should be noted that the interferer power can be much greater than the signal power.

An example of an interference source (ie, narrowband interference) is given with reference to FIG.
Fig. 3b discloses a block diagram of the components of the radio interface of the receiver according to the second mode of realization of the present invention.

  The wireless interface 205 includes a synchronization module 351. The synchronization module 351 is responsible for synchronizing the DFT module 350 of the interface 205 with respect to the received symbols.

The DFT module 350 converts the received symbol from the time domain to the frequency domain, and converts the received symbol y _k into the frequency domain. Here, k represents a carrier index. Received symbols are obtained by converting received single carrier signals into received symbols.

The received symbol in the frequency domain is
y _k = h _k x _k + ν _k
May be represented by Where h _k is the channel response for the carrier with index k and ν _k is the additive noise at the same frequency. The term ν _k is an addition of additive white Gaussian noise (AWGN) noise, such as thermal noise and narrowband interference sources. For narrowband interferers, the dispersion of ν _k is frequency dependent and is represented by σ _k ² .

  Received symbols in the frequency domain are provided to adaptive signal and thermal noise power estimation module 355, threshold-based interference estimation module 353, and provided to adaptive threshold determination and rounding module 358.

  Channel equalization is based on, for example, pilot symbols.

Adaptive signal and thermal noise power estimation module 355 provides a signal and thermal noise power estimate (denoted as P _s ) to threshold-based interference estimation module 353.

Adaptive signal and thermal noise power estimation module 355 provides an estimate of signal and thermal noise power (denoted P _s ) to adaptive threshold determination and rounding module 358.

The adaptive threshold determination and rounding module 358 determines another threshold T _d (eg, equal to the determined power P _s plus 4 dB).

The threshold T _d is then used by the adaptive threshold determination and rounding module 358 to round the received symbol amplitude in the frequency domain according to the following rules:

However, i is a square root of "-1" in both above-mentioned formulas.

  The received symbols processed by adaptive threshold determination and rounding module 358 are then provided to channel estimation module 352 and equalization module 356.

For example, the threshold-based interference estimation module 353 estimates the coarse interference source variance ω _k used by the equalization module for each carrier at index k. This dispersion is calculated from the received power p _k, the threshold T _a as calculated from the signal and the thermal noise power estimation P _s.

For example, T _a = P _s + xdB, and x is 0 to several dB (such as 10 dB at the maximum).

For example, ω _k = max (0, p _k −T _a ).

  This information representing narrowband interference is provided to equalization module 356.

The coarse dispersion ω _k is used by an equalization module 356 that performs equalization in various ways.

  The equalization module 356 may be minimum mean square error equalization (MMSE).

For example, the coarse variance ω _k is used in MMSE equalization as follows:

Where σ ² is an estimate of the thermal noise variance and h _k ^* is the estimated channel for carrier k (provided by the estimation module that performs channel estimation on the received symbols processed by the adaptive threshold determination and rounding module 358) Represents the conjugate of.

Here, it should be noted that σ ² may be set up to a predetermined value.

For example, the coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

  If the equalization is MMSE equalization,

Here, it should be noted that λ _k is a decreasing function of ω _k . For example, λ _k is

Where, for example, a = b = 2P _s .

For example, lambda _k, depending on the value of omega _k, (or, where it is equivalent, depending on the relative values of p _k with respect to the adaptive threshold T _a) 0 and 1 of two Take only the value.

λ _k = 0 corresponds to the positions of the narrowband interferers, and the equalization symbols at these positions are set to null values.

For each carrier k, the output omega _k or lambda _k thresholds based interference estimation module 353 is provided to the equalization module 356.

  The output of the equalization module 356 is provided to the IDFT module 357. The IDFT module 357 may have a different size than the DFT module 350.

  Classical equalization (especially MMSE) assumes complete knowledge of the channel and narrowband interference source dispersion. Narrowband interference variance is easily estimated by the present invention. However, the channel estimation process is sensitive to interference sources. It should be noted that the interferer power can be much greater than the signal power.

  FIG. 4 represents an example of a received signal with narrowband interference.

  A received signal with narrowband interference is represented in the frequency domain. That is, once the DFT module 300 or 350 converts the received symbol from the time domain to the frequency domain.

  The horizontal axis represents frequency and the vertical axis represents the power of the signal received in each frequency band.

  The interference power 43 may be much larger than the signal and thermal noise power 45. If the interference source 41 is a pure sine wave, good reception can be obtained with a signal-to-interference power ratio C / I of -10 dB (ie, an interference source that is 10 times stronger than the signal).

  The present invention is more likely to be a threshold (signals above this threshold are dominant narrowband interference, while signals below this threshold correspond to signal and thermal noise. The purpose is to determine such a threshold (which is more likely).

  If the receiver sets up a threshold according to the received power Pt (denoted 43), the threshold 42 will be much higher than the signal and noise power 45, so the threshold 42 is much less efficient if the interference power is large. Will.

If the signal and thermal noise power 45 (referred to as Ps) is known, a threshold T _a (denoted 44) of about Ps + 4 dB gives good performance. If the receiver Rec knows only the total power Pt, the receiver Rec will have a threshold of about T ′ = Pt + 4 dB (42 and to ensure that the signal is not degraded in the absence of an interference source. Must be used. If the interference source is high, Pt is much larger than Ps and the threshold T′42 is much larger than the optimal threshold T44.

Thus, the present invention estimates Ps before defining the threshold T _a (and optionally the threshold T _d ).

This estimate is based on frequency dependent received power p _k.

To estimate the frequency-dependent reception power p _k, the present invention is block-based (i.e.,
p _k = | y _k | ²
May be implemented).

The present invention may be implemented on an average basis (eg, by applying a filter between (in time) consecutive values of | y _k | ² ):

  Where j is a time block index. A block is a set of samples to which block demodulation (such as DFT, frequency domain processing, IDFT, etc.) is applied.

The a _j value is a smoothing time filter coefficient.

Knowing the p _k value, the estimation of P _s is based on the following principle: “Signal and thermal noise have Gaussian statistics in the frequency domain, while the interference source is“ peak ” And therefore much more sensitive to thresholds. "

  The estimation may be performed iteratively as follows.

The first power estimate of the signal and thermal noise power P _s is the total received power P _t :

be equivalent to.

  Where M is the size of the IDFT executed by the IDFT 307.

The power _pk is then rounded according to a threshold T set up for the current power estimate. For example, T = P ₀ +3 dB = 2P ₀ for the first iteration, and then T = P _i +3 dB = 2P _i .

When p _k ≧ T, p _k = T, and when p _k <T, p _k = p _k .

  After rounding and the average of the rounded power, a correction factor is applied to the average. This correction assumes that the power rounded symbol follows a complex Gaussian law.

The corrected value corresponds to the new power estimate (P _i for the i th iteration).

  An example of the above estimation is given with reference to FIG.

  FIG. 5 discloses a block diagram of the components of a signal and thermal noise power estimation module according to the present invention.

Adaptive signal and thermal noise power estimation module 305 or 355 is provided with a frequency-dependent reception power module 508 to determine the frequency-dependent reception power p _k.

The present invention is block-based, i.e. p _k = | y _k | ²
May be implemented by using

The adaptive signal and thermal noise power estimation module 305 or 355 comprises an average processing module 503. The average processing module 503 averages the total received power P _t to calculate P ₀ . However,

It is.

Power P ₀ is provided to switch 504. Switch 504 provides power P ₀ for the first iteration and provides power P _i when the first iteration is performed. Where i is from 1 to I-1, where I is the total number of iterations performed by the adaptive signal and thermal noise power determination module 305 or 355. The threshold calculation module 505 determines a first threshold (eg, equal to T ₀ = 2P ₀ ) and determines a threshold T _i = 2P _i in subsequent iterations.

A threshold T ₀ (and T _{i in} subsequent iterations) is provided to the rounding module 500. The rounding module 500 rounds all signal power greater than the threshold T ₀ (and T _{i in} subsequent iterations) as follows:
If p _k > T _i, then p _k = T _i .
The value of p _k otherwise is not changed.

  The power value is provided to the average processing module 501. The average processing module 501 averages the power values provided by the rounding module 500.

The average value provided by the average processing module 501 is then provided to the correction module 502. The correction module 502 determines and applies the correction coefficient δ _i .

The correction factor δ ₀ (and δ _{i in} subsequent iterations) is applied to compensate for power loss caused by at least one power rounding to the threshold T ₀ (and T _{i in} subsequent iterations).

  The calculation of the correction factor assumes that the signal with rounded power is a complex Gaussian.

For the i th iteration where i is from 0 to I−1, the correction factor δ _i is determined as follows.

If the signal is a complex Gaussian with power P _i , the power follows the exponential probability law:

It is. However, the parameter

Accompanied by.

When the threshold T _i is applied to the power of such a signal, the average power decreases and the average output power P _AVi is

be equivalent to.

  Therefore,

With respect to P _i, this is

It can be expressed.

  The multiplicative correction term is

be equivalent to.

Since P _AVi and T _i are known, P _i is given by the equation

(For example, by applying a fixed point theorem).

According to a preferred mode of realization of the present invention, a look-up table is used to calculate δ _i from P _AVi and T _i , or directly P _i . In this realization mode, the above equations are used to substitute into the lookup table.

The corrected signal power is provided to the switching module 504. Switching module 504, which provides the threshold calculation module 505 instead of the power _{P 0.}

For example, the number of iterations may be equal to 3-5. The power P _s is equal to the power P _i determined in the last iteration.

  FIG. 6a discloses an example of an algorithm executed by a destination according to the first mode of realization of the present invention.

  The present algorithm is more precisely executed by the processor 200 of the receiver Rec.

  At step S600, the processor 200 commands the synchronization module 301 to synchronize the DFT module 300 with respect to the received symbols. The received symbol is obtained by converting the received single carrier signal into a received symbol.

At next step S601, the processor 200 commands the DFT module 300 to convert the received symbols from the time domain to the frequency domain to become the received symbols y _k in the frequency domain. Here, k represents a carrier index.

Received symbols in the frequency domain may be expressed as:
y _k = h _x x _k + ν _k

At next step S602, the processor 200 commands the signal and thermal noise power estimation module to determine the signal and thermal noise power estimate P _s as disclosed with reference to FIG.

  At step S604, the processor 200 commands the threshold-based interference estimation module 303 to estimate the narrowband interference variance for each carrier.

For example, the threshold-based interference estimation module 303 estimates the coarse interference source variance ω _k used by the equalization module for each carrier at index k. This dispersion is calculated from the received power p _k, the threshold T _a as calculated from the signal and the thermal noise power estimation P _s.

For example, T _a = P _s + xdB, and x is 0 to several dB (such as 10 dB at the maximum).

For example, ω _k = max (0, p _k −T _a ).

For example, the coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

Here, it should be noted that λ _k is a decreasing function of ω _k . For example, λ _k is

Where, for example, a = b = 2P _s .

For example, lambda _k, depending on the value of omega _k, (or, where it is equivalent, depending on the relative values of p _k with respect to the adaptive threshold T _a) 0 and 1 of two Take only the value.

λ _k = 0 corresponds to the position of the narrowband interferer, and the equalization symbols on these carriers are set to null values.

  At step S603, the processor 200 commands the channel estimation module 302 to perform channel estimation based on the received symbols in the frequency domain provided at step S601. Channel estimation module 302 may perform channel estimation based on the pilot symbols.

At step S605, the processor 200 uses the information representing the narrowband interference λ _k or ω _k provided by the threshold based interference estimation for the equalization module 306, using the output of the channel estimation step, and at step S601 Instructs to perform equalization using received symbols in the provided frequency domain.

The coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

  If the equalization is MMSE equalization,

For example, the coarse variance ω _k may be used in MMSE equalization as follows:

  Classical equalization (especially MMSE) assumes complete knowledge of the channel and narrowband interference source dispersion. Narrowband interference variance is easily estimated by the present invention. However, the channel estimation process is sensitive to interference sources. It should be noted that the interferer power can be much greater than the signal power.

  In the next step S606, an IDFT transformation is performed on the sample provided by the equalization step S605.

  FIG. 6b discloses an example of an algorithm executed by a destination according to the second mode of realization of the present invention.

  The present algorithm is more precisely executed by the processor 200 of the receiver Rec.

  At step S650, the processor 200 commands the synchronization module 351 to synchronize the DFT module 350 with respect to the received symbols. The received symbol is obtained by converting the received single carrier signal into a received symbol.

At next step S651, the processor 200 commands the DFT module 350 to convert the received symbols from the time domain to the frequency domain to become the received symbols y _k in the frequency domain. Here, k represents a carrier index.

Received symbols in the frequency domain may be expressed as:
y _k = h _x x _k + ν _k

At next step S652, the processor 200, the signal and the thermal noise power estimation module, commands to determine the signals and thermal noise power estimation P _s as disclosed with reference to FIG.

  At step S653, the processor 200 commands the threshold-based interference estimation module 353 to estimate the narrowband interference variance for each carrier.

For example, the threshold-based interference estimation module 353 estimates the coarse interference source variance ω _k used by the equalization module for each carrier at index k. This dispersion is calculated from the received power p _k, the threshold T _a as calculated from the signal and the thermal noise power estimation P _s.

For example, T _a = P _s + xdB, and x is 0 to several dB (such as 10 dB at the maximum).

For example, ω _k = max (0, p _k −T _a ).

For example, the coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

Here, it should be noted that λ _k is a decreasing function of ω _k . For example, λ _k is

Where, for example, a = b = 2P _s .

For example, lambda _k, depending on the value of omega _k, (or, where it is equivalent, depending on the relative values of p _k with respect to the adaptive threshold T _a) 0 and 1 of two Take only the value.

λ _k = 0 corresponds to the positions of the narrowband interferers, and the equalization symbols at these positions are set to null values.

At step S654, the processor 200 commands the adaptive threshold determination and rounding module 358 to determine another adaptive threshold from the signal and the thermal noise power estimate P _s and perform rounding of the received symbols in the frequency domain.

The adaptive threshold determination and rounding module 358 determines another threshold T _d (eg, equal to the determined power P _s plus 4 dB).

The threshold T _d is then used by the adaptive threshold determination and rounding module 358 to round the received symbol amplitude in the frequency domain according to the following rules:

However, i is a square root of "-1" in both above-mentioned formulas.

  At step S655, the processor 200 commands the channel estimation module 352 to perform channel estimation based on the received symbols processed by the adaptive threshold determination and rounding module 358.

At next step S656, the processor 200 uses the information representing the narrowband interference λ _k or ω _k provided by the threshold-based interference estimation to the equalization module 356, using the output of the channel estimation step and adaptively Command to perform equalization using the received symbols processed in step S654.

The coarse variance ω _k may be used to determine a weighting factor λ _k that is applied to the received symbol y _k in the frequency domain before or after equalization.

  If the equalization is MMSE equalization,

For example, the coarse variance ω _k may be used in MMSE equalization as follows:

  In a next step S657, an IDFT transformation is performed on the sample provided by the equalization step S656.

  Naturally, many modifications can be made to the embodiments of the invention described above without departing from the scope of the present invention.

Claims (14)

A method of canceling narrowband interference in a single carrier signal,
The method comprises the steps performed by the receiver:
-Receiving a single carrier signal and converting the single carrier signal into received symbols;
-Converting the received symbols from time domain to frequency domain to receive symbols in the frequency domain;
-Determining a signal and thermal noise power estimate based on the received symbol power in the frequency domain;
-Estimating the variance of the narrowband interference from the signal and thermal noise power estimation and the received symbol power in the frequency domain;
-Taking into account the estimation of the variance of the narrowband interference, equalizing received symbols in the frequency domain or symbols derived from received symbols in the frequency domain, how to. further,
Determining a weighting factor that depends on the variance of the narrowband interference, wherein the weighting factor is a decreasing function of the variance of the narrowband interference;
The method of claim 1, comprising equalizing received symbols in the frequency domain taking into account the weighting factor.   The method according to claim 2, characterized in that the weighting factor is equal to 1 or a null value, which depends on the estimated variance of the narrowband interference. further,
-Determining the first threshold used for the estimation of the variance of the narrowband interference based on the signal and thermal noise power,
The method as described in any one of Claims 1-3. As a further step,
-Estimating the frequency dependent received power of received symbols in the frequency domain;
-Iteratively determining the signal and thermal noise power from the estimated frequency dependent received power;
The method according to claim 1, comprising: The adaptive signal and thermal noise power determined iteratively are:
Performing a first averaging process of the total received power of the received symbols in the frequency domain;
-In a first iteration, determining a temporary threshold based on said average total received power;
Rounding all received symbol power in the frequency domain greater than the temporal threshold determined in the first iteration;
Performing a second averaging of the rounded power;
-Correcting the second average by a correction factor;
-Determining a subsequent temporal threshold based on the corrected average in subsequent iterations;
-Rounding all power greater than the subsequent temporary threshold;
-Performing a third averaging of the rounded power;
-Correcting the third average by a correction factor;
6. The method of claim 5, wherein the method is determined by performing the temporary threshold determination, the rounding, the third averaging process and the correction a predetermined number of times.   The method of claim 6, wherein the correction factor is calculated assuming that the symbol rounded off the power follows a complex Gaussian law.   The method according to claim 6 or 7, wherein the correction factor is determined using a look-up table.   9. The method according to any one of claims 1 to 8, further comprising performing channel estimation based on received symbols in the frequency domain. As a further step,
-Determining a second threshold based on said signal and thermal noise power estimate;
Rounding the received symbol amplitude in the frequency domain at the determined second threshold, wherein the rounded received symbol in the frequency domain is a symbol derived from the received symbol; Rounding the amplitude of the received symbols in the method.   The method according to claim 10, characterized in that the channel estimation is performed on the rounded received symbols in the frequency domain.   The method according to claim 1, wherein the single carrier signal is a single carrier orthogonal frequency division multiplex modulation signal. An apparatus for canceling narrowband interference in a single carrier signal,
The device is included in a receiver;
The device is
-Means for receiving a single carrier signal and converting said single carrier signal into received symbols;
-Means for converting said received symbols from time domain to frequency domain to be received symbols in the frequency domain;
-Means for determining a signal and thermal noise power estimate based on the received symbol power in the frequency domain;
Means for estimating the variance of the narrowband interference from the signal and thermal noise power estimation and the received symbol power in the frequency domain;
-Means for equalizing received symbols in the frequency domain or symbols derived from received symbols in the frequency domain, taking into account the estimation of the variance of the narrowband interference Do the equipment. A computer program that can be loaded directly into a programmable device,
Computer program comprising code portions or instructions for performing the steps of the method according to any of claims 1 to 12, when the computer program is executed on a programmable device.

Images